Featured in the newest Dialogue Magazine »

Of all the scientific disciplines which resist solution, earthquake prediction certainly ranks close to the top of the list. One need only review the history of recent major events to discover how difficult this issue is.

Consider, for example, the 7.9 magnitude earthquake in Sichuan province in central China, which took place on May 12, 2008. Concerning this event, a commentator in Nature (May 14, 2009) declared: “More so than other quakes, this one has uncovered gaps in earthquake hazard research, both in China and elsewhere.” (p. 153). It happened this way. Beginning in October 2006, the Chinese government established a network of 300 broadband seismometer stations in the western part of Sichuan province. Spaced on average 5 – 30 km apart, these solar powered receptors received signals from an area covering about 370,000 square kilometers. It was the densest array of such sensors in the world, and it definitely was the envy of Western scientists. Nevertheless, when the quake occurred near the towns of Yingxiu and Bailu, it took everyone, including the scientists, by surprise.

It so happens that scientists had formerly studied the nearby Beichuan fault. The data pointed to a very quiet region in terms of seismic activity. And prior to the major event of May 2008, no increase of tremors was recorded. So how could the scientists have foreseen this event? Obviously they could not. Not surprisingly, scientists tend to focus their attention on faults that show active motion, with a history of past large events. They pay attention to regions which might produce a significant event at least every few hundred years, not ones which might be expected every several thousand years!

The Sichuan earthquake of May 2008, then was a surprise because the region had exhibited little recent seismic activity. From this situation, scientists concluded that they had paid too much attention to regions of recent seismic activity and they had therefore underestimated the potential hazard of other areas. But this is a conclusion of futility since what other clues could scientists use to predict earthquake hazard anywhere on the globe?

The Chinese earthquake is an example of a major earthquake event which occurs within a “tectonic plate.” According to current theory, the globe is broken up into a number of large plates which interact with each other at their edges, moving under an adjacent plate edge, or over it. Theory holds that most violent events such as volcanic eruptions and earthquakes will occur at these plate boundaries. However plate tectonic theory gives “no insight into where and when quakes will occur within plates because the interiors of ideal plates should not deform.” (Nature November 5, 2009 p. 87). It is evident therefore that scientists have very little understanding of why earthquakes happen at all inside plate interiors and they certainly are not in a position to predict any such events.

One of the most studied regions of intraplate (within plate) activity is the town of New Madrid, Missouri which lies 2000 km from the nearest plate boundary. In this region however on December 16, 1811 and for three months afterward, a series of strong earthquakes occurred which shook the entire eastern half of the country. At the same time, widespread sand blows occurred. These phenomena result when a quake of magnitude 7.6 or stronger, causes saturated sandy sediment to become liquefied. It then violently erupts from the surface and is deposited over the nearby landscape. The sand from the 1811-12 events, covered wide areas of farmland in a layer so deep that cultivation was difficult for many years to come. Geological analysis of the area, moreover, suggests that there were three or four such major events within the past two thousand years.

The mystery of the events near New Madrid lies in the fact that GPS measurements for the past twenty years show no detectable activity in the landscape. There is no obvious difference between the New Madrid region and the rest of the central and eastern United States. So why have there been such events near New Madrid? Nobody knows, and expert opinion is divided on whether more such events can be expected in that region or not. Even if these events are a thing of the past, that does not explain why major quakes ever happened there in the first place.

If earthquake prediction is difficult in intraplate regions of the world, it actually is not that much better near plate boundaries. The disastrous magnitude 9.0 Japanese earthquake of March 11, 2011 is a case in point. Already in 1978 and still in April 2006, according to National Geographic, the Japanese government had identified the Tokai region 160 km southwest of Tokyo as the probable site of the next great earthquake in Japan (p. 139). Contingency plans for such an event, with minute attention to detail, have been prepared for the area. Sadly, the terrible (worst ever) earthquake when it did come, was near Sendai, 270 km northeast of Tokyo instead of to the southwest.

Similarly the frightful Sumatra-Andaman magnitude 9.1-9.2 earthquake of December 26, 2004, took everyone by surprise. While scientists had equipment deployed and were monitoring various regions of concern around Indonesia, the actual massive event took place in a region with few sensors because practically nobody expected such a major event where it actually happened. (Nature March 2, 2006 p. 31). This event was the first giant event in which even some sensors were able to record the progress of the event. Apparently, over the space of 8 minutes, a rupture occurred which progressed over 1300 km so that a block about 150 km wide was moved about 20 m. The event was much larger than might have been expect in the area. Thus scientists concluded that any such fault can potentially produce a large event. So much for informed assessment of specific risk!

Thus much as we would like to think that we can control risk, our understanding of these events is minimal at best. Meanwhile, societies and governments must take what prudent precautions are possible, such as strict building codes in areas presumed to be at risk. And when such terrible events do strike, we must show compassion and generous assistance to the victims.


Margaret Helder
June 2011

Subscribe to Dialogue